Requirements and design options for Requirements and design options for a vertex detector (Si tracker)a vertex detector (Si tracker)

Requirements and design options for Requirements and design options for a vertex detector (Si tracker)a vertex detector (Si tracker)

Near Detector Workshop, CERN, 31 July 2011

Paul Soler

Near Detector workshop, CERN, 31 July 2011

2

Currently there are three possible techniques for detecting taus (that I know of):– Observation of decay kink of tau: emulsion technique (already

discussed by P. Migliozzi) like in OPERA or CHORUS– Kinematic technique: used by NOMAD to identify taus through the

kinematic analysis of the tau decay– Reconstruction of the impact parameter with a dedicated vertex

detector (ie. Silicon detector, prototyped by NOMAD-STAR)

Pasquale has already covered the emulsion technique I think that a combination of the two other techniques can be

used at a near detector of a neutrino factory.

Tau detection techniques Tau detection techniques

Near Detector workshop, CERN, 31 July 2011

3

Impact parameter detection of taus Impact parameter detection of taus This was invented by Gomez Cadenas et al.

– NAUSICAA: (Si vertex detector): NIM A 378 (1996), 196-220– ESTAR (Emulsion-Silicon Target): NIM A 381 (1996), 223-235

Identify tau by impact parameter (for one prong decay of ) and double vertex (for 3 prong decays)

d~90m

Near Detector workshop, CERN, 31 July 2011

4

NAUSICAA NAUSICAA NAUSICAA proposal:

– Impact parameter resolution -CC=28m

– Impact parameter resolution -CC=62m

Efficiencies:– : =10%– n=10%– n: =23%– Total eff:

=0.85x10%+0.15x23%=12%

– =0.29

Sensitivity at Main Injector with 2x107 CC interactions (MINSIS):

67

106.329.012.0102

48.2)(

oscP

Intrinsic limit tau production Ds:

– 3.5x10-6 at 450 GeV – 1.3x10-6 at 350 GeV– 9.6x10-8 at 120 GeV

PRD 55, (1997),1297. NIMA 385 (1997), 91.

Near Detector workshop, CERN, 31 July 2011

5

NOMAD-STARNOMAD-STAR R&D in NOMAD for short baseline detector based on silicon:

NOMAD-STAR (NIMA 413 (1998), 17; NIMA 419 (1998), 1; NIMA 486 (2002), 639; NIMA 506 (2003), 217.)

Total mass: 45 kg of B4C target (largest density 2.49 g cm-3 for lowest X0=21.7 cm.)

Near Detector workshop, CERN, 31 July 2011

6

NOMAD-STARNOMAD-STAR Aim of NOMAD-STAR: reconstruct short lived particles in a neutrino beam

to determine capabilities detection: use impact parameter signature of charm decays to mimic

impact parameter ~ 62 m, normal charged current (CC) interactions ~30 m

Near Detector workshop, CERN, 31 July 2011

7

Longest silicon microstrip detector ladders ever built: 72cm, 12 detectors, S/N=16:1

Hamamatsu Si Detectors: 300 m thick, 25 m pitch, 50 m readout (640 channels) – total 32,000 channels

VA1 readout: 3 s shaping

NOMAD-STARNOMAD-STAR

1990s technology: could do much better with LHCtype technology and readout

Near Detector workshop, CERN, 31 July 2011

8

CC event

Primary vertex

Secondary vertex

NOMAD-STARNOMAD-STAR

Near Detector workshop, CERN, 31 July 2011

9

Vertex resolution: y = 19 m Impact parameter resolution: 33 m

Double vertex resolution: 18 m from Ks reconstruction

Pull:~1.02

x~33 m

NOMAD-STARNOMAD-STAR

x~18 m z~280 m

Near Detector workshop, CERN, 31 July 2011

10

Charm event reconstruction:– Implementation of Kalman filter – Constrained fit method: extract charm

signatures for each charm mass

Used NOMAD-STAR to search for charm events: marginal statistical accuracy, but was a good proof of principle

NOMAD-STARNOMAD-STAR

Meson Num events

Back- ground

Production Rate

Efficiency

D0 24 13.3 13.3±1.0% 3.5%

D+ 10 3.8 3.8±0.5% 3.5%

Ds+ 11 5.2 5.2±0.6% 12.7%

Total 45 22.3 7.2±2.4%

Near Detector workshop, CERN, 31 July 2011

11

Charm measurement at Near Charm measurement at Near DetectorDetector

Amongst other things, we need to measure charm cross-section to validate size of charm background in wrong-sign muon signature

Charm cross-section and branching fractions poorly known, especially close to threshold

Semiconductor vertex detector only viable option in high intensity environment (~109 CC events/year)

Very hard for emulsion and liquid argon TPC since rate so high

CHORUS 2008

Near Detector workshop, CERN, 31 July 2011

12

Near Detector design will need three sections:– Vertex detector for charm/tau measurement at the front. – High resolution section (SciFi tracker or TRT tracker) for

leptonic flux measurement – Mini-MIND detector for flux and muon measurement

Near Detector Block DiagramNear Detector Block Diagram

beam3 m

3 m

B>1 T

~20 m

High Res DetectorHigh Res DetectorMini-MINDMini-MIND

VertexVertexDetectorDetector

Near Detector workshop, CERN, 31 July 2011

13

Vertex detector:

– Assume: 150x150x2cm3 per layer of B4C (2.49 g/cm3) and 18 layers: total mass = 2.02 tonnes

– Assume 20 layers of silicon: 45 m2 of silicon – Assume silicon strip detectors but could be pixels– About 64,000 channels per layer: 1.28 million channels

Vertex detector dimensionsVertex detector dimensions

beam3 m

3 m

B>1 T

~20 m

High Res DetectorHigh Res DetectorMini-MINDMini-MIND

VertexVertexDetectorDetector

Near Detector workshop, CERN, 31 July 2011

14

Rates: approx 109 CC events/year in 1 ton detector Assume 12% eff from NAUSICAA Inclusive charm production at 27 GeV (from CHORUS):

6.4±1.0%– From 10-30 GeV: ~4%4x107 charm events!!– Charm produced in CC reaction with d or s quark so always

have lepton– Associated charm in NC interaction (see charm review

De Lellis et al. Phys Rep 399 (2004), 227-320) For tau measurement we would use the same analysis

– We would do a search for → and search for -

– However, there is anti-e background that can create D- which looks like signal, but need to identify e+ to reject background!!!

Charm and tau measurementsCharm and tau measurements

So, very important to have light detector with B-field (ie Scintillating Fibre tracker) behind vertex detector to identify positron with high efficiency: – We would have to reject positron background using a similar analysis to that used for

MIND: use a mixture of isolation and kinematic techniques to reject background (ie. In MIND we rejected e background at 10-6 but this was in iron. Assuming no background:

Very tough: probably overestimate in efficiency since background rejection will affect it 8

9101.7

29.012.010

48.2)(

P

Near Detector workshop, CERN, 31 July 2011

15

We can use a vertex detector to identify charm at a NF near detector

Use impact parameter for for one prong events and vertex detection for two prong events

Efficiencies and proof of principle demonstrated in NOMAD-STAR: successful detection of charm

Can achieve two tons of a passive target with silicon strip detectors (45 m2 of silicon, 1.3M channels): not overly challenging in LHC era

No problems at all to measure charm and charm species More challenging to do tau searches due to charm

background, but need to perform full analysis to determine sensitivities to NSI:

ConclusionsConclusions

?107 8P